Yi-34B-Llama / README.md
chargoddard's picture
Create README.md
e65c456
|
raw
history blame
No virus
311 Bytes
[01-ai/Yi-34B](https://huggingface.co/01-ai/Yi-34B/tree/main) with tensors renamed to match standard Llama modelling code. Model can be loaded without trust_remote_code, but the tokenizer can not.
The branch `llama-tokenizer` uses the Llama tokenizer class as well, but *does not* work with LlamaTokenizerFast.